Marijuana remains the most prevalent illicit substance used by adolescents and the number of adolescents receiving treatment for marijuana abuse more than tripled during the last decade. A small number of clinical trials suggest that family-based and individual interventions have efficacy for treating adolescent substance abuse. However, even with these interventions most adolescents fail to reduce their substance use substantially, thus, there remains much room for improvement of treatment services. The overarching goal of this project is to develop and test novel behavioral treatments to enhance treatment outcome in this important treatment population, and in so doing, learn more about mechanisms of change that have broader implications for addiction science. In our initial Stage IB project "Behavioral Treatment for Adolescent Marijuana Abuse", we created, manualized, and pilot tested a unique contingency-management (CM) intervention that combined abstinence-based voucher incentives with contingency management training for parents. A small randomized, clinical trial provided encouraging results. When added to a commonly used cognitive-behavior therapy, CM improved rates of sustained abstinence during treatment. Adolescents receiving this intervention were less likely to relapse over the 9-month follow-up period, however this finding was not as robust as the observed during treatment effects, most likely due to the small sample size and associated low power to detect effects. Despite strong indicators of the efficacy of this CM intervention, there remained room for improvement in increasing rates of treatment response and reducing rates of relapse. Hypothesized mediators and moderators of change indicated that changes in parenting had direct effects on post-treatment marijuana abstinence outcomes, and that abstinence early in treatment was a robust predictor of the CM treatment effect. This proposal will systematically replicate and extend these findings. A Stage II trial will compare three treatment conditions: (1) cognitive behavior therapy (CBT only); (2) CBT plus CM; and (3) CBT plus an enhanced CM model targeting increased early abstinence rates, parenting skills, and maintenance of effects. Replicating the initial demonstration of the positive effects of CM will extend the scientific evidence for use of CM to increase treatment efficacy for substance-abusing adolescents. Testing an enhanced CM model will determine if modifications that are consistent with the underlying behavioral principles and empiricism supporting CM interventions can result in improved outcomes. Last, assessment of potential mechanisms of action, particularly parenting, adolescent psychopathology and impulsivity, will provide scientific information directly relevant to future development of more effective intervention and prevention models of adolescent substance abuse, and will inform us about fundamental mechanisms operating in drug-dependence.